Multi-dimensional data fusion analysis system and method for water pollution traceability

Abstract

The invention discloses a multi-dimensional data fusion analysis system and method for water pollution traceability, and relates to the technical field of water pollution traceability, the investigation range is greatly reduced through a hierarchical response mechanism: the investigation range is locked through the maximum inscribed circle of a cross region, and graphical data display is combined, so that the water pollution traceability is improved. For example, polygon comparison replaces pure numerical value comparison, the abnormal point location recognition speed is remarkably improved, efficient conversion from'global investigation 'to'accurate focusing' is achieved, the feature point location with the most concentrated pollution data is accurately captured through comparison of a reference polygon and an exploration polygon in combination with screening of the same polygon and locking of the feature polygon, and the accuracy of the abnormal point location recognition is improved. The misjudgment risk of a single index is reduced; the feature indexes are further matched with a discharge port index library, and a pollution transmission path is verified through dynamic calculation of'time-flow velocity-distance ', so that the locking of the suspected discharge port is ensured to conform to pollution feature association and meet a physical migration rule, and the traceability deviation is effectively reduced.

Description

Technical Field

[0001] This invention relates to the field of water pollution source tracing technology, specifically to a multi-dimensional data fusion analysis system and method for water pollution source tracing. Background Technology

[0002] Application CN118883880B discloses a water pollution source tracing analysis method and system based on multidimensional data fusion analysis. The method includes: constructing a monitoring network based on the discharge paths of rainwater and sewage to obtain real-time monitoring data; determining whether the real-time monitoring data is abnormal; if an abnormality occurs, issuing an alarm for the abnormal monitoring node; upon receiving the alarm information, performing a fusion analysis of the pollution sources of rainwater and sewage based on a pollution source tracing model based on discharge characteristics analysis and a pollution source tracing model based on three-dimensional fluorescence; positively verifying the results obtained from the fusion analysis through a verification model to obtain the source tracing analysis results; conducting on-site source tracing and evidence collection based on the source tracing analysis results; and predicting the impact of the alarm points based on a rainwater management model and taking emergency measures.

[0003] In the current field of water pollution control, accurate and efficient source tracing is a core prerequisite for targeted pollution control. However, traditional source tracing methods still face many technical bottlenecks. Regarding efficiency, existing technologies largely rely on a combination of fixed-point monitoring and traditional methods. Often, after discovering water quality exceeding standards, a comprehensive upstream investigation is required, leading to problems such as delayed response and blindly expanding the investigation scope. This is especially problematic in areas with complex river branches, where manual judgment of pollution diffusion paths is extremely inefficient. In terms of accuracy, traditional methods often rely on single water quality indicators or isolated monitoring data for analysis. These methods are easily affected by natural background fluctuations and the superposition of multiple pollution sources, making it difficult to establish a reliable correlation between pollution characteristics and emission sources. This results in frequent misjudgments and omissions, and lacks quantitative verification methods for pollution transmission time and paths, leading to insufficient scientific rigor in source tracing conclusions.

[0004] In practical applications, existing technologies are poorly adapted to complex scenarios: on the one hand, traditional methods of manual sampling and numerical comparison are labor-intensive and difficult to implement in remote rivers, dangerous waters, and other areas; on the other hand, when faced with multiple pollution sources, there is a lack of effective data fusion and feature extraction methods, making it difficult to distinguish between primary and secondary pollution contributions, resulting in limited support for law enforcement supervision and governance decisions from the source tracing results; in addition, existing technologies have not formed a standardized, closed-loop system for the entire process, with data fragmented at each stage from the discovery of exceedances to source identification, making it difficult to achieve efficient linkage between "pollution signals - survey data - source location", which restricts the transformation of water pollution source tracing from experience-driven to data-driven. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a multi-dimensional data fusion analysis system and method for water pollution source tracing, solving the problem of insufficient comprehensiveness in water pollution source tracing.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a multi-dimensional data fusion analysis method for tracing water pollution sources, comprising the following steps:

[0007] Step 1: Confirm the exceedance of monitoring data associated with the monitoring node. Based on the confirmation result, determine whether it is necessary to dispatch unmanned vessels to upstream branch nodes for pollution confirmation. If so, confirm the upstream branch nodes associated with the current monitoring node from the water system distribution map and directly dispatch unmanned vessels to monitor pollution data. If not, no dispatch is required.

[0008] Step 2: After the unmanned vessel reaches the upstream branch node location, identify the intersection area of ​​the corresponding branch node from the water system distribution map, and based on the outer contour of the intersection area, identify the largest inscribed circle associated with it, and use the locked largest inscribed circle as the survey range.

[0009] Step 3: Based on the survey area associated with the unmanned vessel, using the survey data at the center point of the survey area as the baseline data, further data surveys are conducted within the survey area. Anomalies are identified during the data survey process, thereby pinpointing the associated abnormal river channels.

[0010] Based on the number of survey indicators set inside the unmanned vessel, generate a corresponding number of polygons, confirm the connections between the polygons and their corner points, bind each connection to a single survey indicator, and assign the indicator parameters associated with the unit length of the corresponding connection.

[0011] Control the unmanned vessel to reach the center point of the survey area and conduct data survey. Based on the surveyed data and the constructed polygon, confirm the data points associated with the corresponding lines, and connect the data points on adjacent lines to confirm the reference polygon associated with the survey data corresponding to the center point.

[0012] Then, the unmanned vessel is controlled to survey the points on the circle of the survey area in sequence, and the surveyed polygon associated with the corresponding point is confirmed according to the same construction method as the reference polygon.

[0013] The center points of the surveyed polygon and the reference polygon are then aligned sequentially. The survey indices corresponding to the aligned lines are all the same. The survey is then identified to determine whether there is a set of more than one set of corner points that exceed those of the reference polygon. If so, the surveyed polygon is marked as a polygon to be determined and the corner points that exceed the reference polygon are recorded. If not, no marking is performed.

[0014] Confirm the survey indicators associated with the corner points recorded for each undetermined polygon, and record undetermined polygons with the same survey indicators as equal polygons. Then confirm the area parameters of equal polygons and lock the undetermined polygon with the largest area parameter as the characteristic polygon of the corresponding equal polygon.

[0015] Identify whether the set of survey indicators associated with the corresponding feature polygon exists in the set of survey indicators of other feature polygons. If it exists, remove this feature polygon and retain the other feature polygons. If it does not exist, do not perform any processing.

[0016] The points on the circle of the survey area with the feature polygon are recorded as feature points. The direction of movement from the center point to the feature points is taken as the feature direction. Then, the specific flow direction of different rivers is confirmed from the water system distribution map. The angle between the specific flow direction and the feature direction is confirmed. The river with the smallest angle data is recorded as an abnormal river.

[0017] Step 4: Confirm the discharge outlets associated with the abnormal river channel upstream, then, based on the confirmed abnormal indicators, screen out potential discharge outlets from the associated outlets, and finally, based on water flow velocity and relevant time, identify and display suspected discharge outlets:

[0018] Identify the discharge outlets associated with the upstream of the abnormal river from the river system distribution map, simultaneously identify several sets of discharge indicators associated with the corresponding discharge outlets, and then identify the abnormal indicators associated with the feature points. Discharge outlets that satisfy the condition that the survey indicators associated with the abnormal indicators are all within several sets of discharge indicators are recorded as undetermined discharge outlets.

[0019] The monitoring process associated with the abnormal node is confirmed. The moment when the water pollution signal is determined is recorded as the initial moment. Based on the monitoring process, the time point when the corresponding monitoring data reaches its peak is confirmed. The time length between the initial moment and the time point is locked and recorded as the relevant time T.

[0020] The travel routes associated with different pending discharge outlets and abnormal nodes are confirmed in sequence. The route lengths of different rivers are confirmed from the travel routes. The water flow velocity measured in different rivers is then confirmed. The travel time is confirmed based on the route lengths of different rivers and the water flow velocity. Several sets of travel times are summed to confirm the total time characteristics.

[0021] Undetermined emission outlets that satisfy |total time characteristic - T| ≤ Y1 are recorded as suspected emission outlets and displayed directly, where Y1 is a preset value; otherwise, no calibration is performed.

[0022] A multi-dimensional data fusion and analysis system for tracing water pollution sources includes:

[0023] The monitoring unit is set up at a designated location in the waterway to confirm whether the associated monitoring data exceeds the standard. If any parameter exceeds the standard, a water pollution signal is generated directly.

[0024] The control center, based on the generated book pollution signal, identifies the upstream branch node from the water system distribution map, dispatches the unmanned vessel, and simultaneously controls the unmanned vessel to conduct on-site surveys.

[0025] At the survey range confirmation end, after the unmanned vessel reaches the upstream branch node position, it confirms the intersection area of ​​the corresponding branch node from the water system distribution map, and locks the survey range based on the range characteristics of the intersection area.

[0026] The abnormal river channel confirmation end uses the survey data of the center point of the survey range associated with the unmanned vessel as the benchmark data, and then conducts data surveys based on the survey range. From the data survey process, abnormal points are confirmed, thereby locking down the associated abnormal river channels.

[0027] The suspected discharge outlet confirmation end identifies the discharge outlets associated with the upstream of the identified abnormal river channel, and filters out potential discharge outlets from the associated discharge outlets based on the identified abnormal indicators. Then, based on the monitoring progress, it confirms the relevant time, and finally locks down and displays the suspected discharge outlets based on the flow velocity and relevant time.

[0028] This invention provides a multi-dimensional data fusion analysis system and method for tracing the source of water pollution. Compared with existing technologies, it has the following advantages:

[0029] This invention significantly reduces the scope of investigation through a hierarchical response mechanism: Step 1 quickly triggers unmanned vessel surveys of upstream branch nodes based on the out-of-range signals of monitoring nodes, avoiding the blindness of traditional manual inspections; Step 2 locks the survey range by the maximum inscribed circle of the intersection area, and combines graphical data display (such as polygon comparison) to replace simple numerical comparison, which significantly improves the speed of anomaly point identification and realizes the efficient transformation from "full-area investigation" to "precise focusing";

[0030] The reliability of the results was enhanced by multi-dimensional data association verification: Step 3, by comparing the baseline polygon with the surveyed polygon, combined with the screening of equal polygons and the locking of feature polygons, accurately captured the feature points where the pollution data was most concentrated, reducing the risk of misjudgment by a single indicator; Step 4 further matched the feature indicators with the emission outlet indicator library, and verified the pollution transmission path through dynamic calculation of "time-flow velocity-distance", ensuring that the locking of suspected emission outlets not only conforms to the pollution feature association, but also meets the physical migration law, effectively reducing the source tracing deviation. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0032] Figure 2 This is a schematic diagram of the principle framework of the present invention. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] First Embodiment

[0035] Please see Figure 1 This application provides a multidimensional data fusion analysis method for water pollution source tracing, including the following steps:

[0036] Step 1: Confirm the exceedance of monitoring data associated with the monitoring nodes. Based on the confirmation results, determine whether it is necessary to dispatch unmanned vessels to upstream branch nodes for pollution confirmation. Specifically, several different monitoring nodes are set up in the river area. Each monitoring node can monitor multiple sets of water quality indicators. If the water quality indicators exceed the standards, it indicates that there is water pollution. Source tracing is required to identify the pollution source. Each different water quality indicator has different set standards, which are all formulated in advance by relevant personnel. The water quality indicators generally include: pH value, COD (chemical oxygen demand), ammonia nitrogen, heavy metals (such as lead and mercury), organic matter (such as pesticides and antibiotics), etc.

[0037] The specific method for confirming that the standard has been exceeded is as follows:

[0038] The monitoring nodes compare different monitoring data with preset standards. If a monitoring data exceeds the associated standard, a water pollution signal is generated directly, and the monitoring node is recorded as an abnormal node.

[0039] Based on water pollution signals, the upstream branch nodes associated with the current monitoring node are identified from the water system distribution map (the so-called branch nodes are different river confluences, and their monitoring nodes are generally set in the main river channel). The water system distribution map is a preset graphic, which is prepared in advance by relevant personnel according to the layout of the water system. All branch nodes are marked in advance in the water system distribution map. Based on the confirmed upstream branch nodes, unmanned vessels are dispatched to the upstream branch nodes to monitor pollution data.

[0040] The unmanned vessel is equipped with the same monitoring sensors, and the monitoring method is consistent with that of the corresponding monitoring node to monitor water quality.

[0041] Step 2: After the unmanned vessel reaches the upstream branch node location, confirm the intersection area of ​​the corresponding branch node from the water system distribution map, and based on the range characteristics of the intersection area, lock in the survey range. The process of locking in the survey range is as follows:

[0042] The specific flow direction of each river is determined from the river distribution map. The river is extended according to its boundary line, and the intersection area generated by the corresponding extension area of ​​multiple rivers is confirmed. (The river has a corresponding boundary line. By extending the boundary line in the river distribution map, the extension area associated with the corresponding river can be confirmed. When each river is extended, the intersection area associated with the corresponding extension area can be confirmed.)

[0043] Based on the outer contour of the intersection area, the largest inscribed circle belonging to this intersection area is identified, and the identified largest inscribed circle is taken as the survey range of the unmanned vessel. Specifically, from the outer contour of the intersection area, the center point is identified first. The center point can be identified by using a two-dimensional coordinate system and the mean value of the coordinates. The mean value of the mean value is the corresponding center point. Then, the lines connecting different contour points are identified from the outer contour. The connecting lines pass through the center point. The smallest connecting line is selected from several connecting lines. The smallest connecting line is the diameter of the corresponding largest inscribed circle, which can quickly lock the survey range associated with the corresponding unmanned vessel.

[0044] Step 3: Based on the survey range associated with the unmanned vessel, and using the survey data at the center point of the survey range as the baseline data, further data surveys are conducted within the survey range. During the data survey process, anomalies are identified, thereby pinpointing the associated abnormal river channels. Specifically, in this survey processing process, a graphical display method is used to quickly identify points exceeding the standard. Compared to numerical comparison, this processing method is faster. Only the specific measurement standard needs to be confirmed, and the abnormal river channels can be quickly located based on the corresponding survey data, facilitating subsequent water pollution source tracing and enabling an adaptive source tracing process.

[0045] The specific method for locking down abnormal river channels is as follows:

[0046] Based on the number of survey indicators set inside the unmanned vessel, generate the corresponding number of polygons, confirm the lines connecting the polygons and their corner points, bind each line to the corresponding single survey indicator, and assign the indicator parameters associated with the unit length of the corresponding line (the so-called unit length is not limited here, it only represents one unit length, which is convenient for the subsequent actual survey process).

[0047] Control the unmanned vessel to reach the center point of the survey area and conduct data survey. Based on the surveyed data and the constructed polygon, confirm the data points associated with the corresponding lines (based on the surveyed index parameters and the measurement standard of the corresponding lines, confirm the position of the points on the corresponding lines that belong to this survey. If the position of the point exceeds the corresponding line, the corresponding line can be extended). Connect the data points on adjacent lines to confirm the reference polygon associated with the survey data corresponding to the center point.

[0048] Then, the unmanned vessel is controlled to survey the points on the circle of the survey area in sequence, and the surveyed polygon associated with the corresponding point is confirmed according to the same construction method as the reference polygon.

[0049] The center points of the surveyed polygon and the reference polygon are then aligned sequentially, with the survey indices corresponding to the aligned lines being identical. The system identifies whether there is a set of corner points of the surveyed polygon that exceed those of the reference polygon. If so, the surveyed polygon is marked as a polygon to be determined, and the corner points that exceed the reference polygon are recorded. If not, no marking is performed. For example, if a surveyed polygon has an A index that exceeds the A index of the reference polygon, but other indices do not exceed the A index, then this surveyed polygon will be marked as a polygon to be determined. If multiple sets of indices exceed the A index, it will also be marked as a polygon to be determined.

[0050] Identify the survey indicators associated with the corner points of each undetermined polygon, and record undetermined polygons with the same survey indicators as equal polygons. Then, confirm the area parameters of the equal polygons and lock the undetermined polygon with the largest area parameter as the characteristic polygon of the corresponding equal polygon. For example, if there are two undetermined polygons, both of which exceed index A, then the two surveyed polygons are equal polygons. Then, confirm the area parameters, and the undetermined polygon with the largest area parameter is the characteristic polygon of the corresponding equal polygon.

[0051] The system identifies whether the set of survey indicators associated with the corresponding feature polygon exists in the set of survey indicators of other feature polygons. If they do, the feature polygon is removed and other feature polygons are retained. If they do not exist, no processing is performed (depending on the specific confirmation process, there is generally only one set of feature polygons. If there are two pollution sources, there will be two sets). For example, if feature polygon o1 exceeds indicator A, feature polygon o2 exceeds indicators A and B, and feature polygon o3 exceeds indicators A, B, and C, then survey polygons o1 and o2 need to be removed, and only feature polygon o3 needs to be retained. In the actual survey process, the survey point associated with o3 is the point with the most concentrated pollution data. Based on the confirmed most concentrated point, the relevant confirmation of the river channel can be carried out.

[0052] The points on the circle of the survey area with the feature polygon are recorded as feature points. The direction of movement from the center point to the feature point is taken as the feature direction. Then, the specific flow direction of different rivers is confirmed from the water system distribution map. The angle between the specific flow direction and the feature direction is confirmed. The river with the smallest angle is recorded as an abnormal river. If there is no angle, the feature direction is extended in the opposite direction to confirm the angle. After the abnormal river is confirmed, the abnormal indicators associated with the corresponding feature points can be used to trace upstream and identify various discharge outlets associated with the upstream of the abnormal river. The location of the discharge outlets whose corresponding discharge indicators are consistent with the abnormal indicators is confirmed. Then, according to the data climbing process of the corresponding monitoring nodes in step one, the suspected discharge outlets are identified and confirmed, and the source is traced and displayed.

[0053] Step 4: Based on the confirmed abnormal river channels, identify the associated discharge outlets upstream of the corresponding abnormal river channels. Then, based on the confirmed abnormal indicators, screen out potential discharge outlets from the associated outlets. Next, based on the monitoring progress, confirm the relevant time periods, and then, based on flow velocity and relevant time periods, pinpoint and display the suspected discharge outlets. The process of confirming suspected discharge outlets includes:

[0054] Identify the discharge outlets associated with the upstream of abnormal river channels from the river system distribution map, simultaneously identify several sets of discharge indicators associated with the corresponding discharge outlets, and then identify the abnormal indicators associated with the characteristic points (that is, the survey indicators associated with the corresponding corner points). Discharge outlets that satisfy the condition that all the survey indicators associated with the abnormal indicators exist within several sets of discharge indicators are recorded as undetermined discharge outlets. For example, if the indicators discharged by a certain discharge outlet are A, B, C, D and E, and its abnormal indicators are A, B and C, then the abnormal indicators all exist within the identified discharge indicators, and the corresponding discharge outlet is an undetermined discharge outlet.

[0055] The monitoring process associated with the abnormal node is confirmed. The moment when the water pollution signal is determined is recorded as the initial moment. Based on the monitoring process, the time point when the corresponding monitoring data reaches its peak is confirmed. The time length between the initial moment and the time point is locked and recorded as the relevant time T.

[0056] The travel routes associated with different pending discharge outlets and abnormal nodes are confirmed sequentially (which can be confirmed from the water system distribution map). The route lengths of different rivers are confirmed from the travel routes, and then the water flow velocities measured in different rivers are confirmed (generally real-time speeds; under normal circumstances, water flow velocities do not change drastically over time, but fluctuate within a certain range). Based on the route lengths of different rivers and the water flow velocities, the travel time is confirmed. Several sets of travel times are summed to confirm the total time characteristics. The calculation methods for length and flow velocity are all described here, so they will not be elaborated on one by one. Because the rivers are wide enough, if the corresponding water flow is not increased exponentially (which can be caused by torrential rain, so it is not possible to measure during the rainy season), the measured water flow velocities will not differ significantly.

[0057] Undetermined discharge outlets that satisfy the condition |Total Time Characteristic - T| ≤ Y1 are recorded as suspected discharge outlets and displayed directly, where Y1 is a preset value. Otherwise, no calibration is performed. The specific value is determined by the operator based on experience, and is generally set to 30 minutes. This value is a maximum value, and under normal circumstances, the confirmed error will not exceed 10 minutes. In the actual survey and processing process, surveys cannot be carried out on rainy days, because rainwater will seriously affect the survey indicators and the survey flow rate, which will lead to errors in the survey results.

[0058] Based on the confirmed suspected emission outlets, relevant personnel can then effectively trace the source of pollution, significantly reducing the difficulty of tracing and ensuring the accuracy of the tracing process.

[0059] Second Embodiment

[0060] Combination Figure 2 A multi-dimensional data fusion analysis system for tracing water pollution sources includes:

[0061] The monitoring unit is set up at a designated location in the waterway to confirm whether the associated monitoring data exceeds the standard. If any parameter exceeds the standard, a water pollution signal is generated directly.

[0062] The control center, based on the generated book pollution signal, identifies the upstream branch node from the water system distribution map, dispatches the unmanned vessel, and simultaneously controls the unmanned vessel to conduct on-site surveys.

[0063] After the unmanned vessel reaches the upstream branch node, the survey range confirmation end identifies the intersection area of ​​the corresponding branch node from the water system distribution map. Based on the range characteristics of the intersection area, the survey range is locked. Specifically, the upstream branch node has a designated location node, marked in the corresponding water system distribution map. When the control center controls the unmanned vessel to reach the designated location, a feedback signal is generated. Subsequently, based on this feedback signal, the survey range confirmation end automatically generates the corresponding survey range and feeds it back to the control center. The control center then controls the unmanned vessel to conduct on-site surveys based on the survey range and transmits the associated survey data to the abnormal river channel confirmation end. During the on-site survey, the survey data of the center point of the survey range is confirmed first, and then the survey data associated with other points on the corresponding circle are confirmed in turn.

[0064] The abnormal river channel confirmation end uses the survey data of the center point of the survey range associated with the unmanned vessel as the benchmark data, and then conducts data surveys based on the survey range. From the data survey process, abnormal points are confirmed, thereby locking down the associated abnormal river channels.

[0065] The suspected discharge outlet confirmation end identifies the discharge outlets associated with the upstream of the identified abnormal river channel, and filters out potential discharge outlets from the associated discharge outlets based on the identified abnormal indicators. Then, based on the monitoring progress, it confirms the relevant time, and finally locks down and displays the suspected discharge outlets based on the flow velocity and relevant time.

[0066] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0067] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A multidimensional data fusion analysis method for tracing the source of water pollution, characterized in that, Includes the following steps: Step 1: Confirm the exceedance of monitoring data associated with the monitoring node. Based on the confirmation results, determine whether it is necessary to dispatch unmanned vessels to the upstream branch node for pollution confirmation. Step 2: After the unmanned vessel reaches the upstream branch node location, identify the intersection area of ​​the corresponding branch node from the water system distribution map, and lock the survey range based on the range characteristics of the intersection area. Step 3: Based on the survey range associated with the unmanned vessel, take the survey data of the center point of the survey range as the baseline data, and then conduct data surveys based on the survey range. From the data survey process, identify anomalies and thus locate the associated abnormal river channels. Step 4: Based on the confirmed abnormal river channel, identify the associated discharge outlets upstream of the corresponding abnormal river channel, and based on the confirmed abnormal indicators, screen out potential discharge outlets from the associated discharge outlets. Then, based on the monitoring progress, confirm the relevant time, and then lock and display the suspected discharge outlets based on the water flow velocity and the relevant time.

2. The multidimensional data fusion analysis method for water pollution source tracing according to claim 1, characterized in that, In step one, the specific method for confirming that the monitoring data exceeds the standard is as follows: The monitoring nodes compare different monitoring data with preset standards. If a monitoring data exceeds the associated standard, a water pollution signal is generated directly, and the monitoring node is recorded as an abnormal node. Based on water pollution signals, the upstream branch nodes associated with the current monitoring node are identified from the water system distribution map. The water system distribution map is a preset graphic, and the branch nodes are all marked in advance within the water system distribution map. Based on the identified upstream branch nodes, unmanned vessels are dispatched to the upstream branch nodes to monitor pollution data.

3. The multidimensional data fusion analysis method for water pollution source tracing according to claim 1, characterized in that, In step two, the process of locking in the survey area is as follows: The specific flow direction of each river is determined from the water system distribution map, and the river is extended according to its boundary line. The intersection areas generated by the extension areas of multiple rivers are also identified. Based on the outer contour of the intersection area, the largest inscribed circle belonging to this intersection area is identified, and the identified largest inscribed circle is taken as the survey range of the unmanned vessel.

4. The multidimensional data fusion analysis method for water pollution source tracing according to claim 1, characterized in that, In step three, the specific method for conducting data surveys based on the survey area is as follows: Based on the number of survey indicators set inside the unmanned vessel, generate a corresponding number of polygons, confirm the connections between the polygons and their corner points, bind each connection to a single survey indicator, and assign the indicator parameters associated with the unit length of the corresponding connection. Control the unmanned vessel to reach the center point of the survey area and conduct data survey. Based on the surveyed data and the constructed polygon, confirm the data points associated with the corresponding lines, and connect the data points on adjacent lines to confirm the reference polygon associated with the survey data corresponding to the center point. Then, the unmanned vessel is controlled to survey the points on the circle of the survey area in sequence, and the surveyed polygon associated with the corresponding point is confirmed according to the same construction method as the reference polygon. The center points of the surveyed polygon and the reference polygon are then aligned sequentially. The survey indices corresponding to the aligned lines are all the same. The survey is then checked to see if there is a set of more than one set of corner points that exceed those of the reference polygon. If so, the surveyed polygon is marked as a polygon to be determined and the corner points that exceed the reference polygon are recorded. If not, no marking is performed.

5. A multidimensional data fusion analysis method for tracing water pollution sources according to claim 4, characterized in that, In step three, the specific method for locking the abnormal river channel is as follows: Confirm the survey indicators associated with the corner points recorded for each undetermined polygon, and record undetermined polygons with the same survey indicators as equal polygons. Then confirm the area parameters of equal polygons and lock the undetermined polygon with the largest area parameter as the characteristic polygon of the corresponding equal polygon. Identify whether the set of survey indicators associated with the corresponding feature polygon exists in the set of survey indicators of other feature polygons. If it exists, remove this feature polygon and retain the other feature polygons. If it does not exist, do not perform any processing. The points on the circle of the survey area with the characteristic polygon are recorded as characteristic points. The direction of movement from the center point to the characteristic point is taken as the characteristic direction. Then, the specific flow direction of different rivers is confirmed from the water system distribution map. The angle between the specific flow direction and the characteristic direction is confirmed. The river with the smallest angle data is recorded as an abnormal river.

6. The multidimensional data fusion analysis method for water pollution source tracing according to claim 1, characterized in that, In step four, the specific method for identifying suspected emission outlets is as follows: Identify the discharge outlets associated with the upstream of the abnormal river from the river system distribution map, simultaneously identify several sets of discharge indicators associated with the corresponding discharge outlets, and then identify the abnormal indicators associated with the feature points. Discharge outlets that satisfy the condition that the survey indicators associated with the abnormal indicators are all within several sets of discharge indicators are recorded as undetermined discharge outlets. The monitoring process associated with the abnormal node is confirmed. The moment when the water pollution signal is determined is recorded as the initial moment. Based on the monitoring process, the time point when the corresponding monitoring data reaches its peak is confirmed. The time length between the initial moment and the time point is locked and recorded as the relevant time T. The travel routes associated with different pending discharge outlets and abnormal nodes are confirmed in sequence. The route lengths of different rivers are confirmed from the travel routes. The water flow velocity measured in different rivers is then confirmed. The travel time is confirmed based on the route lengths of different rivers and the water flow velocity. Several sets of travel times are summed to confirm the total time characteristics. Undetermined emission outlets that satisfy the condition |total time characteristic - T| ≤ Y1 are recorded as suspected emission outlets and displayed directly, where Y1 is a preset value.

7. A multidimensional data fusion analysis method for tracing water pollution sources according to claim 6, characterized in that, No calibration is performed on undetermined emission outlets that do not meet the condition that |total time characteristic - T| ≤ Y1.

8. A multidimensional data fusion analysis system for water pollution source tracing, the system operating according to any one of claims 1-7, characterized in that, include: The monitoring unit is set up at a designated location in the waterway to confirm whether the associated monitoring data exceeds the standard. If any parameter exceeds the standard, a water pollution signal is generated directly. The control center, based on the generated book pollution signal, identifies the upstream branch node from the water system distribution map, dispatches the unmanned vessel, and simultaneously controls the unmanned vessel to conduct on-site surveys. At the survey range confirmation end, after the unmanned vessel reaches the upstream branch node position, it confirms the intersection area of ​​the corresponding branch node from the water system distribution map, and locks the survey range based on the range characteristics of the intersection area. The abnormal river channel confirmation end uses the survey data of the center point of the survey range associated with the unmanned vessel as the benchmark data, and then conducts data surveys based on the survey range. From the data survey process, abnormal points are confirmed, thereby locking down the associated abnormal river channels. The suspected discharge outlet confirmation end identifies the discharge outlets associated with the upstream of the identified abnormal river channel, and filters out potential discharge outlets from the associated discharge outlets based on the identified abnormal indicators. Then, based on the monitoring progress, it confirms the relevant time, and finally locks down and displays the suspected discharge outlets based on the flow velocity and relevant time.

Images